The overall goal of this research is to determine the usefulness of applying concepts and procedures developed in cognitive psychology to the study of how people learn-by-example. To attain this goal we propose to investigate how instructions to encode a model's performance in different ways (e.g., vivid visual imagery or trait impressions) interact with rate-of-exposure to the performance; b) how long it takes for subjects to find different types and numbers of targets in a sequence; c) how instructions to differentially encode a model's behavior affect the way in which people divide it into "meaningful units"; d) how these units affect other perceptual phenomena; e) how long it takes people to "search" their immediate and long-term memory of a model's behavior to determine whether or not different types of targets were present in the sequence; and f) how sets, behavioral unit size and number of categories into which the behavoral units are placed affect long-term retention of the information contained in a model's performance. These areas represent an attempt to divide the cognitive processes governing learning-by- example into three major areas: initial processing, working memory and long-term memory. Of primary concern is how different types of information contained in a model's performance (e.g., his specific motor responses, the changes he produces in his environment, his personality traits, etc) are processed at each of these points.